Handwritten Digit Recognition using ANN (MNIST)

This project implements an Artificial Neural Network (ANN) to classify handwritten digits from the MNIST dataset. It demonstrates a simple feedforward architecture using fully connected layers.

🧠 Model Architecture

• Input Layer: 784 neurons (28x28 pixels)
• Hidden Layers: 2 Dense layers with ReLU activation
• Output Layer: 10 neurons with Softmax (for 10 digit classes)
• Loss Function: Categorical Crossentropy
• Optimizer: Adam

📊 Results

• Achieved ~97% accuracy on test data
• Visualized predictions on sample images
• Trained and evaluated with minimal preprocessing

🛠️ Technologies Used

• Python
• TensorFlow / Keras
• NumPy
• Matplotlib
• Jupyter Notebook

📁 Files

Jupyter notebook with complete code Sample output images from predictions

🔗 Dataset

MNIST dataset is publicly available via Keras datasets:
https://keras.io/api/datasets/mnist/